A review of the outcomes of past attempts at establishing sustainable seed producer groups in Nepal showed that after donor support was withdrawn a lack of marketing skills resulted in the groups no longer producing seed. Learning from this review, when we initiated new attempts at establishing sustainable seed producer groups in Chitwan district, Nepal, we emphasized the strengthening of their marketing and managerial capabilities rather than training in technical issues such as seed quality control. We imparted marketing skills to committee members of farmer groups at an initial training course in Chitwan in 2001. This inspired at least three existing farmer groups in Chitwan, already established for other agricultural activities, to enter into cereal and legume seed production and its marketing. Following their establishment in 2002 we supported them initially by purchasing some of their seed production. This was progressively withdrawn and, after three years, the groups independently marketed all of their substantial seed production. They built up capital reserves mainly from subsidies and by attracting funds from new shareholders with only a small contribution from retained profits that were only about 5% of total turnover. The capital reserves reduced or eliminated the need for loans thus increasing the chances that the enterprises would be sustainable. In contrast, other government-supported groups had practically no cash reserves despite substantial seed sales. By 2010, two of the three groups were still operating and had substantially increased turnover. Shareholders who were also seed producers benefited from being members of the group and from an increased income of 10% by producing seed instead of grain. Our intention in supporting these groups was to promote the scaling out of new rice varieties produced by client-oriented breeding (COB) or identified by participatory varietal selection but most of the seed that was produced was of obsolete varieties. Policies are needed to preferentially promote new varieties by supplying more information about them and increasing the subsidy on their seeds compared with older varieties. Continuing promotion by the organizations that bred them is also desirable but constrained by limited funding for COB.

Puddling coarse-textured soils for rice culture in the irrigated tract of the Indo-Gangetic Plains causes high soil strength in the upper layers. This may adversely affect growth and yield of following upland crops. It is possible that no-tillage (NT) in wheat (without residues of preceding rice crop) could aggravate this problem and reduce fertilizer nitrogen (N) use efficiency. In certain production scenarios, NT has been reported to be advantageous because it allows for earlier planting of wheat by eliminating delays caused by tillage. This study examined the combined effects of two crop establishment options for wheat in relation to fertilizer N and planting time following puddled rice cultivation in an irrigated environment of Punjab, northwest India. Combinations of two establishment systems, NT-direct planting and conventional-tillage (CT) with soil disruption to 0.10 m depth in main plots, with two N rates, 120 and 150 kg ha−1 in subplots, were evaluated. Variation in planting time, 31 October (D1) and 7 and 10 November (D2), was used to evaluate effect of planting-earliness. Under D1 in NT, grain yield of wheat was comparable to that under D2 in CT. However, under no advancement of wheat planting in NT, grain yield was 0.2–0.3 t ha−1 less than that in CT. This yield reduction in NT could be overcome by adding 30 kg ha−1 more fertilizer N suggesting that tillage enhanced N use efficiency. These tillage gains are ascribed to the greater extraction of profile stored and applied water and nutrients because of denser crop rooting caused by reduction in soil strength and less weed competition. Higher N use efficiency in CT suggests that farmers could achieve fertilizer N savings with this system. Comparing the tillage systems showed that NT was more cost-effective than CT even after accounting for the cost of additional fertilizer N (saving of Rs. 1685 ha−1) to overcome associated yield penalties under no advancement in planting time.

Using farmer resource typologies, adaptability analysis and an on-farm mother and baby trial approach, we evaluated the production risks of alternative maize-legume crop combinations for smallholder farmers in Chisepo, central Malawi between 1998 and 2002. Production benefits and risks of four soil fertility and food legumes, pigeonpea (Cajanus cajan), groundnut (Arachis hypogaea), tephrosia (Tephrosia vogelii) and mucuna (Mucuna pruriens), intercropped or rotated with maize, were compared by 32 farmers in 4 farmer resource groups (RGs) of different wealth status. The calculation of lower confidence limits was used to determine the production risk of the crops. Alternative crop technologies presented different risks to farmers of different wealth status, and the degree of risk affected their choice of soil fertility management strategy. The better-resourced farmers (RG 1) had larger yields with all crop combinations than the poorly resourced farmers (RG 4). Legumes integrated with maize significantly (p < 0.001) raised maize grain yields by between 0.5 t ha−1 and 3.4 t ha−1, when compared with sole crop unfertilized maize. Fertilized maize was less of a risk for the better-resourced farmers (RG 1 and RG 2), and it yielded well when combined with the legumes. Maize-legume intercrops yielded more and were associated with less risk than the maize-legume rotations. Maize intercropped with pigeonpea was predicted overall to be the least risky technology for all RGs. We conclude that new crop technologies may pose more risk to poorly resourced farmers than to wealthier farmers.

Informal markets receive little attention from governments and researchers, despite their centrality to farmers' seed security. This paper documents the importance of informal markets for supplying seed and restocking critical plant genetic resources in normal and stress periods. It analyses farmers' rationales for using such markets and their strategic actions in selecting out seed from grain. Conceptual aids for differentiating among market goods – grain, ‘implicit seed’ and seed – are presented, including tracing of agro-ecological seed sources, traders’ seed management behaviour and seed/grain price patterns. Ethiopian case material gives rare insight into how different scales of traders manage the seed/grain divide. Better understanding of informal markets is an important precursor to strengthening them as such markets have unrealized potential to deliver more and higher quality seed, and a greater range of modern and local varieties. Support for informal seed markets could usefully feature in rural livelihood and social protection programmes, but this will require basic shifts in interventions and further refinements in market analysis.

A study was conducted in Suba area, central highlands of Ethiopia, to assess the net return, land and labour productivity, and the return to scale of cereal farming practice. Seventy-five farmers belonging to three local wealth classes (poor, medium and rich) were randomly selected and interviewed about inputs and outputs related to cereal farming for the production year 2007/2008. Farm soil properties were investigated to check the variability in soil quality among the wealth classes. Benefit:cost ratio (BCR), net returns and annual profit were used to indicate the worthiness of the cereal farming activity. The return to scale was estimated by using the Cobb–Douglas production function. The results show that cereal farming is a rewarding practice, with the rich households gaining more profit than the poor. Farm size was the most important variable that affects the net return. There is an increasing return to scale. However, it is unlikely that farmers will have more land than they own at present because of the land shortage problem in the country caused by the increasing human population. Thus, attention should be given to minimizing the costs of production through proper regulation of domestic fertilizer costs and increasing labour productivity especially for the poor and medium households. The use of manure and compost as an additional fertilizer should also be promoted.

Due to escalating costs of imported fertilizers, there is renewed interest in the use of local nutrient resources in managing soil fertility in Kenya. We tested the effect of two organic materials, farmyard manure (FYM) and Tithonia diversifolia (tithonia), and an inorganic N fertilizer, urea, when applied alone or in combination with three inorganic P sources, triple superphosphate (TSP), Minjingu phosphate rock (MPR) and Busumbu phosphate rock (BPR), on maize yields and financial benefits. The study was conducted for three consecutive seasons, from March 2007 to August 2008 in western Kenya. FYM and tithonia were applied to supply 20 kg P ha−1 in treatments where they were used either alone or in combination with the inorganic P sources while 40 kg P ha−1 was from the inorganic P sources in the combination. Where urea was used, the inorganic P sources were applied at 60 kg P ha−1. When applied in combination with urea, MPR was a better P source for maize than TSP or BPR. However, when applied in combination with FYM or tithonia, TSP was the best P source. Treatments including tithonia were more effective in increasing maize yields than those without it with a similar total P application rate. The agronomic effectiveness of tithonia did not, however, translate to economic attractiveness, mainly due to very high labour costs associated with its use. FYM when applied alone at 20 kg P ha−1 was the only treatment that exceeded a benefit:cost ratio of 2 and, therefore, the most likely, of the tested technologies to be adopted by farmers.

Integrated crop-livestock farming in the Guinea savanna of West Africa is often assumed to lead to synergies between crop and livestock production, thereby improving the overall productivity and resilience of agricultural production. Whether these synergies actually occur remains poorly studied. On-farm trials were conducted in northern Nigeria over a period of four years to assess the agronomic and economic performance of maize-legume systems with and without the integration of livestock (goats). Groundnut-maize rotations with livestock achieved the highest carry-over of nutrients as manure from one season to the next, covering approximately one-third of the expected N, P and K uptake by maize and reducing the demand for synthetic fertilizers. However, the advantage of lower fertilizer costs in rotations with livestock was offset by higher labour costs for manure application and slightly lower values of maize grain. Overall, no clear agronomic or economic benefits for crop production were observed from the combined application of manure and synthetic fertilizer over the application of synthetic fertilizer only, probably because the amounts of manure applied were relatively small. Legume-maize rotations achieved higher cereal yields, a better response to labour and fertilizer inputs, and a higher profitability than maize-based systems with no or only a small legume component, irrespective of the presence of livestock. Livestock at or near the farm could nevertheless make legume cultivation economically more attractive by increasing the value of legume haulms. The results suggested that factors other than crop benefits, e.g. livestock providing tangible and non-tangible benefits and opportunities for animal traction, could be important drivers for the ongoing integration of crop and livestock production in the savanna.

Agricultural research and development (R&D) would benefit from reliable yet cheap technology uptake indicators to guide decision making. The paper explores the use of village surveys to monitor technology use and illustrates this through two empirical case studies into tillage dynamics in the Trans-Gangetic Plains in northwest India. The first case study is a revisit of 50 communities surveyed earlier in Haryana State. The second case study is a new and wider representative sample of 120 villages across Haryana and Punjab States. The case studies illustrate that after an initial rapid spread of tractor-drawn zero tillage drills for wheat seeding in these intensive systems, the zero + reduced tillage area seems to have stabilized there at between a fifth and a quarter of the wheat area. Conventional tillage for wheat continues to decline, with an increased use of rotavators making up the difference – but its intensive shallow tillage goes against the conservation agriculture tenets. The paper illustrates the potential of village surveys to provide timely and cost-effective feedback to agricultural R&D.

Phosphorus (P) is one of the main limiting nutrients in the semi-arid regions where pearl millet is grown; its deficiency leads to poor seedling establishment and eventually poor crop yield. Experiments were carried out in pots and field-like conditions to evaluate the effect of seed priming and seed coating with P on the shoot biomass at two and four weeks after sowing (WAS), and on the panicle and stover yield at maturity of three hybrid varieties of pearl millet in low P Alfisol. Overall, seed priming did not increase shoot biomass at two and four WAS. In pots, seed coating at a rate of approximately 400 g P ha−1 increased vegetative biomass over 400% at early stages, and panicle yield by about 50%, over the non-coated treatment, with genotypic variation in the magnitude of the response. In field-like conditions, seed coating restored stover biomass to 85–100% of that in the non-limiting P treatment, whereas the panicle yield remained 25–35% lower than in the non-limiting P treatment, but still 45–65% higher than the non-coated treatment. P deficiency delayed the time to flowering by 20–24 days compared to the non-limiting P treatment, but plants in the seed coating treatment reduced that gap by 10–14 days. Seed P coating treatment appears a valid option to promote pearl millet seedling establishment and then to boost yield under low soil P conditions. The maintenance of an effect of seed coating on panicle yield in pots suggests a positive early effect on developmental processes before panicle initiation.

Leucaena leucocephala is cultivated at close spacings that do not permit intercropping. This has been a discouraging factor for small landholders who need regular income to establish leucaena plantations and benefit from the rapidly expanding market for wood. Therefore, on-farm experiments were conducted near Bhadrachalam, Khammam district, Andhra Pradesh, India, from August 2001 to January 2006, to study the effect of reducing tree density and modifying tree geometry on the growth of leucaena and productivity of intercrops. The inter-row spacing of 1.3 m in farmers' practice was increased up to 13 m to examine whether wide-row planting and grouping of certain rows would facilitate extended intercropping without sacrificing wood yield. Tree density treatments tried were 1.3 × 1.3 m, 3 × 0.75 m, 3 × 1 m, 5 × 0.8 m and 3 × 2 m which gives densities of 5919, 4444, 3333, 2500 and 1666 trees ha−1, respectively. Tree geometry treatments tested were 7 × 1 m paired row spacing (7 × 1 PR), 10 × 1 m triple row spacing (10 × 1 TR), and 13 × 1 m four rows (13 × 1 FR) with a constant tree population of 2500 trees ha−1. Cowpea (Vigna unguiculata) was the intercrop. While changes in tree density affected diameter at breast height (DBH) significantly, modification of tree geometry did not affect tree height and DBH. Marketable wood and dry biomass productivity was highest with 3 × 0.75 m spacing, and reducing tree density and alteration of tree geometry reduced the biomass considerably. In 2001, 2002 and 2003 seasons, respectively, tree spacing at 3 m produced mean yields of 97, 23 and 11% of the sole crop cowpea yield whereas modified tree geometry treatments produced mean yields of 97, 61 and 20% of sole crop yield. The widest spacing (13 × 1 FR) recorded 95, 73 and 30% of the sole crop yields during 2001, 2002 and 2003, respectively. Net returns from intercropping of leucaena in 3 × 0.75 m spacing was 36% higher than that of the farmers' practice. Although wider tree geometry treatments recorded lower net returns, they provided higher intercrop yields and returns in the first two years of plantation establishment. Therefore, it can be concluded that in regions where annual rainfall is around 1000 mm, leucaena can be planted at a spacing of 3 × 0.75 m for improving intercrop performance, higher tree productivity and returns.

Impact assessment hinges on reliable indicators of technology uptake, but this poses particular challenges for new agronomic practices in emerging economies. This paper presents a triangulation approach to assess technology diffusion and illustrates it by assessing the extent and dynamics of zero tillage wheat in South Asia's rice-wheat systems. Various diffusion indicators from three different data methods (secondary data, supply-side surveys, demand-side surveys) are contrasted for two different locations (India's Haryana State and Pakistan's Punjab province) for the period 2000–2004. Such multi-indicator, multi-site and multi-year triangulation enhances the robustness of the inferences and provides the necessary perspective. The three sources coincide in terms of diffusion of zero tillage wheat having increased since 2000, but in particular the empirical surveys show a much slower uptake and subsequent stagnation in the Pakistan Punjab study area. The paper thereby reiterates the need for empirical ground truthing of technology uptake and the need for robust and complementary diffusion indicators.

Many food production systems in sub-Saharan Africa are constrained by phosphorus (P). We hypothesized that within legume-cereal rotation systems: targeting P to the legume phase leads to higher system productivity, and that use of grain legumes leads to better economic returns than use of herbaceous legumes. Four P application regimes: (i) no P, (ii) P applied every season, (iii) P applied in season 1 only and (iv) P applied in season 2 only were tested for four seasons in three cropping systems (continuous maize, mucuna-maize rotation and soybean-maize rotation) in a split plot experiment set up in Nyabeda, western Kenya. Treatments where P was applied were better than no P treatments. While continuous cereal systems showed the need for application of P every second season, rotation systems involving mucuna and soyabean indicated that application in one out of three seasons could be sufficient. Nitrogen fertilizer equivalence was 52 to >90 kg N ha−1 for soyabean and 37 to >90 kg N ha−1 for mucuna, depending on P fertilization and season. Analysis of marginal rates of return (MRR) showed that soybean-maize rotation with one application of P was the most economically viable option, with an MRR of at least 147% compared to other non-dominated options.

Lack of effective chilling during the dormant season is one of the major problems when apples are growing under a tropical climate. We evaluated the response of different apple cultivars (Golden Delicious, Gala, Fuji, Granny Smith and Jonagold) grown on M9 rootstock with different dormancy-management practices. The trials were carried out between 2004 and 2006 in a tropical mountain area (Tigray, Ethiopia), where chilling conditions are poor with the aim of improving and synchronizing the bud break and the blossoming period of these apple cultivars. Two-year-old well-feathered trees were planted in two experimental trial sites in a randomized complete block design. Trees were subjected to the following treatments in two sets of experiments: one defoliation per year only; two defoliations per year, one defoliation followed by 1% hydrogen cyanamide (Dormex) treatment; one defoliation followed by 2% Dormex treatment; one defoliation followed by 4% winter oil; one defoliation followed by 0.5% Dormex and 2% winter oil; and a control with no defoliation or dormancy breaking treatments. The results show positive effects of the dormancy breaking agents on the productivity of the trees after defoliation, with comparable results for the effectiveness of both Dormex and winter oil. There were no statistically significant differences between the Dormex doses. The defoliation treatment alone was not sufficient to break dormancy for the cultivars Golden Delicious, Granny Smith or Gala but showed promising results with dormancy breaking on Jonagold. Yields increased as a result of better flowering time synchronization within a tree but even with the dormancy treatments the length of the flowering period was still spread over five weeks, where under a more temperate climate it lasted two to three weeks. The average fruit weight of Jonagold and Granny Smith can be considered as a good fruit quality while the fruit of other diploid cultivars like Golden, Gala and Fuji were rather small, which indicates that fruit thinning by hand will be a necessity for these cultivars. Red colouration of the apples on the cultivars Gala and Jonagold was excellent and meets the standards necessary for commercialization of these fruits. The sugar concentration of the fruits and the fruit firmness at harvest was high. The results of these first trials indicate that it is possible to develop new apple production in the mountain region of Tigray, Ethiopia.

Conservation agriculture (CA) systems are based on minimal soil disturbance, crop residue retention and crop rotation. Although the capacity of rotations to break pest and disease cycles is generally recognized, other benefits of crop rotations in CA systems are seldom acknowledged and little understood. We monitored different conventional and CA cropping systems over the period from 2005 to 2009 in a multi-seasonal trial in Monze, southern Zambia. Both monocropped maize and different maize rotations including cotton and the green manure cover crop sunnhemp (Crotalaria juncea) were compared under CA conditions, with the aim of elucidating the effects of crop rotations on soil quality, soil moisture relations and maize productivity. Infiltration, a sensitive indicator of soil quality, was significantly lower on conventionally ploughed plots in all cropping seasons compared to CA plots. Higher water infiltration rate led to greater soil moisture content in CA maize treatments seeded after cotton. Earthworm populations, total carbon and aggregate stability were also significantly higher on CA plots. Improvements in soil quality resulted in higher rainfall use efficiency and higher maize grain yield on CA plots especially those in a two- or three-year rotation. In the 2007/08 and 2008/2009 season, highest yields were obtained from direct-seeded maize after sunnhemp, which yielded 74% and 136% more than maize in the conventionally ploughed control treatment with a continuous maize crop. Even in a two-year rotation (maize-cotton), without a legume green manure cover crop, 47% and 38% higher maize yields were recorded compared to maize in the conventionally ploughed control in the two years, respectively. This suggests that there are positive effects from crop rotations even in the absence of disease and pest problems. The overall profitability of each system will, however, depend on markets and prices, which will guide the farmer's decision on which, if any, rotation to choose.

The present study documents critical analysis of drought-induced physiological responses in mulberry (Morus spp.) with insights into growth dynamics and leaf productivity. The study was performed for two years in a two-phase experimental design combining both field (experiment no. 1) and glasshouse (experiment no. 2) observations. In field assays, we surveyed 15 mulberry genotypes under two irrigation regimes: well-watered (20 to 24 irrigations in each growing season) and water-limited (irrigated once in a fortnight in each growing season). The genotypes were assessed for variation in key leaf gas exchange characteristics: net photosynthetic rates (Pn), stomatal conductance of CO2 (gs), transpiration rates (E) and instantaneous water use efficiency (WUEi). Leaf yield/plant was considered to determine the tolerance index (TI). Drought stress severely down-regulated leaf-level physiological variables in the susceptible genotypes resulting in poor leaf yield. However, genotypes S-13 and V-1 performed better in terms of leaf gas exchange and proved their superiority over other genotypes in drought tolerance. Conversely, genotypes DD and Bogurai were highly susceptible to drought. Under glasshouse conditions, the combined leaf gas exchange/chlorophyll a fluorescence measurements further dissected out stomatal and non-stomatal restrictions to Pn. As internal/ambient CO2 ratio (Ci/Ca) decreased concurrently with gs in non-irrigated stands, it appeared that greater stomatal limitation to Pn was associated with decreased photo-assimilation and leaf yield production. Further, higher leaf temperature (TL) (>35 °C) and down-regulation of maximum quantum yield of photosystem II (Fv/Fm) were apparent in the susceptible compared to the tolerant genotypes, which indicated chronic photoinhibition due to photo-inactivation of photosystem II centres in the susceptible genotypes. Drought-induced trade-offs in biomass allocation were also highlighted. Overall, our results suggest that greater rooting vigour and leaf hydration status, minimal stomatal inhibition and stabilized photochemistry might play major roles in maintaining higher Pn and associated gas exchange functions in drought-tolerant mulberry genotypes under water stress conditions. The higher leaf yield production in tolerant than susceptible genotypes can be attributed to minimal plasticity in foliar gas exchange traits and better quantitative growth characteristics under low water regimes.

Field irrigation is costly in arid regions, and over-fertilization of farmland can lead to high groundwater nitrate levels and significant environmental challenges. Integrative aquaculture and agriculture (IAA) systems allow the reuse of water and nutrients to offset production costs while promoting greater sustainability. The aim of this study was to test the effectiveness of an IAA system using treatments formed from one water source, groundwater (GRND) or fish pond effluent (EFF), and one chemical fertilizer regime, eliminated (ELIM) or historical (HIST). Treatments were applied to field plots of barley or cotton. There were typically positive effects of EFF applications on crop growth and yield relative to GRND applications under identical fertilizer regimes. However, GRND-HIST almost always outperformed EFF-ELIM, suggesting that substituting effluent irrigations for a historical fertilization regime without pond biosolid or reduced fertilizer applications could be detrimental to crop production.

On-farm nutrient management practices were evaluated in the irrigated rice-wheat cropping system (RWCS) in northwestern India. Large farmers used comparatively higher doses of nutrients than small and medium farmers for both rice and wheat crops. On average, a field is treated with organic manure after every 15 years for rice and 53 years for wheat. Nitrogen use was generally more than the recommended dose for both crops. Phosphorus use in wheat was nearly as per recommendation, while in rice it was below the recommended dose. Potassium use was almost nil; however, the use of Zn (particularly in rice) is becoming common. Use of biofertilizers was negligible. Overall, farmers applied higher doses of organic manures, K, Zn and biofertilizers in rice than wheat, and there was increasing trend in their use with increase in the size of land holdings. Grain yield increased significantly with increase in the size of land holding in rice but remained almost same in wheat. This study generated comprehensive data on nutrient use pattern in the RWCS, thus providing necessary input to researchers, planners and developmental agencies who wish to enhance food security.

The impact of three different pollination treatments (self-, cross-, free-) on the degree of shotberry (seedless fruit) formation of the olive tree cultivars (cvs) Koroneiki, Kalamata, Mastoidis and Amigdalolia was studied for three consecutive years. Controlled crosses were made for the cross-pollination treatments, while for the free pollination treatment flowers were allowed to receive pollen from more than 40 cultivars present in the field. Significant differences were recorded between treatments, cultivars and years. The lowest degree of shotberry formation was observed in free-pollinated trees while the highest was in self-pollinated trees of all cultivars. Low air temperature incidents during the flowering period increased shotberry formation. Cultivars Koroneiki and Mastoidis were reciprocally the most effective pollinator varieties in reducing shotberries. Cultivar Koroneiki could be considered as the most suitable pollinator to reduce shotberries in Kalamata. When cv. Amigdalolia was cross-pollinated by cvs Koroneiki and Mastoidis the degree of shotberry formation was lower when compared to cross-pollination by cv. Kalamata.

With a focus on eastern Africa, the results of research on the ecophysiology and water relations of tea are reviewed. In particular, work undertaken at the Ngwazi Tea Research Station (formerly Unit) in southern Tanzania is synthesized and interpreted in relation to work reported from elsewhere in Africa (and beyond where appropriate). Topics covered include factors influencing: the components of yield, yield distribution, root growth, stomatal behaviour, photosynthesis, transpiration, xylem water potential and how cultivars vary in their responses to water stress. A companion paper (Carr, 2010) reviews the results of research on the irrigation of tea, water productivity, and its practical significance to the tea industry and the wider scientific community.

Aerobic rice is considered a viable agro-technology to cope with the looming crisis of water supply that threatens the sustainability of irrigated rice production systems. Rice adapted to aerobic conditions requires less water than that grown under conventional irrigation management. A field study was conducted at Cuttack, India, during the dry season (January–May) in 2005 and 2006 to determine the critical soil moisture regime at the root-zone depth (30 cm) for sustaining optimum growth and grain yield of aerobic rice variety ‘Apo’ (IR 55423-01). Irrigation at 0, 20 and 40 kPa soil moisture tension resulted in similar grain yields (4.90–5.25 t ha−1 in 2005 and 4.35–4.50 t ha−1 in 2006). The seasonal water requirement in treatments receiving irrigation at 20, 40 and 60 kPa soil moisture tensions was 28.4, 42.8 and 60.7% lower than that at 0 kPa soil moisture tension, but the yield declined significantly at 60 kPa, i.e. by 42.8% in 2005 and 36.7% in 2006. Irrigation at 40 kPa soil moisture tension ensured maximum water productivity of 0.90, 0.47 and 0.53 g grain kg−1 water with respect to evapotranspiration, irrigation plus rainfall and irrigation alone, respectively. Thus, irrigation at 40 kPa soil moisture tension may be considered critical for optimum grain yield and maximum water productivity of aerobic rice in Indian cultivation conditions.